Use of vibrational spectroscopy and chemometrics for rapid screening of ciprofloxacin in freeze-dried milk powder

ABSTRACT The detection of antibiotic residues in dairy products is essential to ensure food safety. This study used freeze-dried milk powder as a controlled laboratory matrix because freeze-drying helps preserve the chemical integrity of milk and enables homogeneous spectral acquisition in solid form. The aim was to evaluate mid-infrared (MIR) and near-infrared (NIR) spectroscopy combined with chemometric methods for the rapid screening of Ciprofloxacin residues. Milk samples were fortified with different antibiotic concentrations, freeze-dried, and analyzed by MIR and NIR spectroscopy. To increase dataset representativeness, synthetic samples were generated using Principal Component Analysis (PCA), preserving the multivariate structure of the original spectra and increasing sample density within the concentration classes. Chemometric analyses included PCA to explore spectral variability and Partial Least Squares Discriminant Analysis (PLS-DA) to classify samples as antibiotic-free, containing Ciprofloxacin up to the maximum residue limit (MRL), or above the MRL. The best performance was obtained using MIR spectroscopy with Savitzky–Golay first-derivative preprocessing, synthetic samples, and four latent variables, reaching average accuracies of 0.994 in calibration and 0.991 in external validation, with sensitivity and precision values above 0.90 for all classes. These findings indicate that the proposed spectroscopic–chemometric approach is a promising rapid, non-destructive screening tool for Ciprofloxacin residues in milk powder and supports further studies with real milk samples.